Moblin: Proof that Corporate Support Needed.
If anything the sudden appearance of Moblin 2.0 Beta and its excellent User Interface has proven, beyond any doubt, that corporate support is essential if linux — and the open source community — is going to survive beyond a very very small niche.
Linux on the server has been doing well despite Microsoft’s pretty good record with Windows Server in the past few years (and contrary to its failings with Vista), its dominant position on the desktop and its proven marketing muscle and the reason for this has been that Linux on the server has had the support of many large corporations living off it.
This has not been the case on the linux desktop, and it is probably the reason why so little has been achieved in the past seven or so years in that field. On one hand, the stagnant Gnome 2 platform barely kept alive primarily by Redhat and Novell that depend on it and on the other the interesting and fresh KDE4 platform that’s extremely immature and incomplete and leaves thousands of everyday use cases unsupported.
The problem is: as weird as it may seem, the gap between the leading desktop environments and linux is widening; sure, linux has moved forward, but so has OS X and so have Windows. What’s more, both leading platforms have now created the foundations for the desktop of the next five or ten years, while at the same time chaos, old and insufficient APIs and poor documentation are the norm in linuxland.
Users and developers alike have gotten spoilt by the ease of use of modern systems to the point where the linux desktop is increasingly looking and feeling dated. Skin-deep eyecandy as that provided by KWin and Compiz is not enough [is never was anyway] and not really helping usability and projects that will take another four years before they can be considered mature just don’t cut it: they’ll be obsolete by then anyway and chances are very few will even use them by that time.
We are now reaching a crucial point in time that will either spell the end of linux desktop as a promising platform [at least under the existing frameworks] or will solidify it as a contender. With Redhat and Novell having officially abandoned the linux desktop as a commercial product for the time being and with the remaining corporate players either too small or too incapable of supporting the fundamental development efforts required to bring desktop linux to the forefront, it is only through individual projects by stellar programmers, small teams in startups and — of course — the support of large corporations outside of the software development ecosystem.
Come to think of it, in the short history of linux, most of the innovation on the linux desktop has come from relatively small companies that staked their existence on it and focused on producing innovative solutions: Eazel, creating Nautilus, Trolltech (now Qt/Nokia) creating Qt, OpenedHand creating Clutter, HelixCode/Ximian creating lots of things off the top of my head. Most of those either failed or were bought by much larger corporations. But the work they did was phenomenal and game changing in many respects. This cannot be said about Canonical, Mandriva or many others claiming the credit for the successes of linux; while there is no doubt that, say, Mandriva’s early efforts helped bring people to linux or that Canonical’s fork of Debian and subsequent packaging efforts really helped during some of the darkest hours of desktop linux, but these were just too small contributions.
Most of the companies that really did help linux were founded, active and bankrupt in the years around the turn of the century, when ’200x is going to be the Year of Linux on the desktop’ still meant more than an anachronistic joke among linux geeks.
The Moblin project is probably one of the very few examples of linux software that could be branded as polished, well thought-out design and high performance. It is a testament as to what is possible when large corporations, such as Intel, support small innovative companies, such as OpenedHand, to create new and innovative products that may end up changing the market where they operate.
While the desktop is probably going through a tough time with more and more people looking to mobile devices for the future and ignoring that no matter how much time elapses, they’ll probably always work sitting at a desk and that the notion of ‘working’ is mutually exclusive to ‘netbooks’ or ‘iPhones’, I’m sure that it’s pretty hard to convince anyone to invest in revamping the linux desktop from the ground up.
Yet Intel’s work with Moblin is filling me with hope that Linux just might be given a new chance on the desktop if larger corporations realise how much they could gain by investing it it now and create the foundations for a long time to come.
Update: I have updated the post by clarifying that it was the appearance of the beta of Moblin 2.0 that was ‘sudden’ and very pleasant, not the project in general. Moblin has been in the works for a while [first based on Ubuntu, then on Fedora], but the UI featured in the beta hasn’t shown its face to the world since now.
Open your eyes…big companies have a big hand in Linux development. Not as corporations but as individuals that happen to work for big corporations. :)
I’m not sure from what perspective you look at it, but I suppose you like Linux for some unknown reasons and still prefer it was something different. Would that be a correct description?
The trouble I see in this is a mish-mash of information, some incoherent and other not fair. Do you seriously say “so little has been achieved in the past seven or so years in that field” or is it a rhetoric aim to make a point? These exaggerations diminish any creative criticism.
Sure, I agree that the Linux desktop has to improve, on the other hand I can’t in your article see any concrete examples about what you’re talking about. What I see though is a flood of similar articles on the Internet with obscure purposes. Some of these articles are of course the product of spin preparing us for the official release of Win 7. Your article isn’t and neither is the author of “Why Desktop Linux sucks”, but you both seem to strive to push Linux into a direction it never was designed for. Somehow I get the impression that these attempts is less based on appreciation of Linux features and more on hostility toward Microsoft as a company, or maybe not hostility but an eager to see a competitor in the image of Microsoft. That if anything will kill off Linux and the whole philosophy behind it!
Moblin is an example of what happens when a company takes care and try to figure out beforehand what has to be accomplished and not simply rush out a half-baked product. Weren’t there other companies before Intel which used Linux on nebooks? Yes there were, but they didn’t listen and didn’t take use of the Linux community to get it right from the start. They didn’t say “hey, we need to before release make sure that we’ve got good support for all huawei modems”, but rushed them out as the-absolute-mobile-but-still-computer-like-device. So in this case we might argue that big companies are capable damaging Linux by ignorant strategies.
It’s easy to a couple of bits and create a story around them. Intel has been a contributor for a long time so it’s not like it happened with Moblin (and Intel has been a bit ambivalent in the development of Moblin). A polished desktop is fine but if all bolts and pieces aren’t in their place before it’s kind of useless. Or is this the Windows approach?
Lastely, Windows is probably fine for those who like it, but Linux is based on another computing idea. Linux won’t die on the desktop because you who don’t like that idea say so, it will only die if those who share this common approach to computing suddenly got extinguished from the surface of earth. Windows and Mac depend on companies in their turn dependent on their own ability to generate money. Their users don’t play a great part of how they develop. Linux develop only in the direction its users want it to and it doesn’t depend on the same companies being around during its life cycle. These systems are different in many ways.
I can compare with how it was a couple of years ago and I tell you Linux has never been as strong at the desktop as now. I’m not talking about market share figures or netbooks, but about attitudes among youths. You’re pretty young but still like me a “breed” who grow up on Microsoft crack (take this with a sense of humour).
@908: You’re absolutely right. And that’s the problem. Corporations should provide more and [most importantly] more focused support, not just paychecks to individuals. Intel did this with Moblin; a hollistic approach as opposed to a distribution of funds to good programmers. Redhat does it to some extent and so does Novell. But most corporations don’t provide wider support for fundamental development and specifications. We need that.
@KimTjik: Hello, and thanks for the comment.
First of all, it’s obvious that I do like linux as perhaps the best known ambassador of what can be achieved with open source; I’m not particularly fond of the kernel [there's too much wrong with it] or parts of the software in the user space [there are numerous examples of pretty bad engineering in GUI platforms on linux, parts of the linker, various end-user applications etc]. Now you write that I’d ‘still prefer it was something different’. This is an invalid assertion that assumes that linux *is* something. Let me spare you the trouble: it isn’t something. It’s not what you, a linux hacker or even linus torvalds. Linux is free technology and based on this technology you can make wonderful things. While i’m sure many can say that linux is already wonderful, I can point to other pieces of software that can achieve the same thing much easier, faster, better; so yes, I’d like to see linux improve on the desktop.
Of course I do. I’ve been using linux since 1998. Not that early, but early enough to remember how different it was to 2002-2003 when GNOME 2.x came out along with KDE 3.x. Step back for a second, ignore that microworld of bugs and marginal improvements to existing applications and paradigms and you’d probably agree with me that practically *nothing* has been done on desktop linux since then, save sustaining an already obsolete codebase [and in the case of KDE4, putting together an alpha quality desktop with incoherent components and design that --- while it shows promise --- is far from complete and will --- at this rate --- take many years to be on par with its competitors in so many critical areas].
That’s probably because you read the article in isolation, while it’s been a favourite past time of mine on this blog. I can’t blame you. There’s literally hundreds of things that are wrong with Linux desktop today though; configuration, usability, coherence, performance, stability, a unified audio subsystem, extremely bad/broken search indexing, mediocre æsthetics, incompatible application packaging and the list could go on forever.
Linux wasn’t designed for *anything* in particular. Linux has evolved in the server space because more people sat down and wrote code to optimize it for the server as they cared about it; whether they were corporations making [or saving] money off it or individuals. People that — most likely — had to respond to comments similar to yours that went kind of like this: ‘why do you bother tinkering with linux. it’s a hobby OS, it has no use in the server room. Use AIX/Irix/Solaris/Whatever.’
So, no: linux wasn’t designed to be used for something in particular. It’s a collection of software that can be used for anything; that’s part of its beauty and the beauty of any open source piece of software.
Not at all. It is exactly because I do appreciate what linux has to offer that I’d like to see it become better; this much should be obvious. There’s no hostility against Microsoft; they make a lot of products, some pretty good, most pretty bad and they own the desktop market and a large part of the low end server market; so, yeah, of course they inevitably get into the picture.
Again: if you think that linux has survived until now by individual, ‘philosophically’ minded people, you’re sorely mistaken; most of the work happens by paid programmers that work for big corporations; until now they have focused on what brings them money: the corporate desktop and the server. I’m arguing that this has to change. Including more work on usability and æsthetics will not take anything away from the philosophy behind linux or the open source community that gave birth to it. On the contrary.
I agree.
Absolutely. But this doesn’t mean that big corporations cannot help linux improve; as it’s clearly the case for the better part of this decade. The linux ecosystem wouldn’t be anywhere on the map without the funding, support and code by: IBM, Novell/Suse, Redhat, Sun and the list goes on. I hope this is clear to you when you try to portray linux as the product of individuals, the mythical sole programmer that writes everything in her bedroom or home office. While there are definitely thousands of extremely talented programmers out there doing amazing work on open source software, corporate support, either via paychecks, donations, or other financial support programmes, is and has been essential for open source and linux in particular for many years.
A usable, polished desktop is essential in my opinion. Right now linux development is in flux; lots of things are changing on the desktop; from the freedesktop.org efforts [DeviceKit etc] to the upcoming GNOME 2.30/3.x to a more mature, non-alpha quality KDE 4.3 and beyond.
So in a way, your argument is flawed; today the linux desktop is neither particularly usable nor does it have solid ‘bolts and pieces’ behind it.
Windows has been a mess for years and they are now starting to fix it; the transition to managed code, the cleaning up of the old, horrible win32 API, the gradual sorting out of the security issues are part of this effort.
Again: No. Linux is based on no computing idea whatsoever. I feel at home using it, as I feel comfortable using any UNIX-like operating system, but that doesn’t mean that linux should be stuck to a 1970s philosophy entrenched by hard core programmers or their wannabe cheerleaders that train themselves to use unintuitive systems and are proud of the esoteric intricacies that they spent so much time learning through man pages and forums. If that’s your idea of what linux is, then you’re probably happy as things are right now, with your fvwm and xterm or even the current GNOME/KDE platforms and whatnot, but you cannot possibly believe that your views are universal. That’s at best naive and at worst arrogant.
Ahh, the Linux Myth. You’re confusing the fact that software is open source with corporate contribution/participation.
It’s true that Windows and Macintosh development is closed and inaccessible to most users; It’s not true at all that linux doesn’t depend on a few, pretty large and established corporations. There are significant barriers to entry to contributing to a large piece of software like the linux kernel, openoffice.org, java or firefox. The fact is that most of the core work on those projects happens by professionals working on the code on behalf of companies. The difference is that in the case of open source, if corporations suddenly disappeared one day, the community could [eventually, after a relatively long time] pickup where they left off and continue developing the code, whereas this would be impossible in the case of closed source software.
The Linux Desktop has gone a long way since the mid 1990s. Its competitors have been moving fast [especially OS X, which I feel should be in many ways a reference point for linux on the desktop] and the gap is widening, either due to stagnation or lack of pace. Things change fast in the world of software and linux desktop platforms may radically improve in the next few years and leapfrog the closed source variants, however I strongly believe that corporate support, like the work by Intel/OpenedHand on Moblin, is essential before this is going to happen.
Thanks a lot for the comment and the constructive criticism.
ignoring that no matter how much time elapses, they’ll probably always work sitting at a desk and that the notion of ‘working’ is mutually exclusive to ‘netbooks’ or ‘iPhones’, I’m sure that it’s pretty hard to convince anyone to invest in revamping the linux desktop from the ground up.
this is great passage ;)
btw, Moblin like any other will fail. Linux means failure open source model doesn’t provide enough money to keep the interesting entity motivated
Hi,
I believe that for corporations to get behind Linux on the desktop it must get easier to deploy closed source software to it. LSB says all distributions should standardize on RPM, but that is quite obviously not happening. I think the RPM from rpm5.org looks most promising personally. Having to deploy an application is a nightmare on Linux. Maybe Adobe and Google get help but even they only ship stuff to a handful of distributions.
I know that reducing choice goes against all that Linux is founded on. On the other hand, I believe that the abundance of choice is difficult to navigate and could be reduced in some ways to add make life easier for VARs. Shuttleworth was hoping to get everyone deploying in synch. For something deployed for 5 to 7 years, it should not matter if you are behind. Synchronize the kernels and libcs and you will be able to reduce the headache enormously. Support costs will go down and security will increase (distributions will be able to share patches). As secure as Linux is touted to be, it could be more secure. Again, LSB is helping by pushing NSS. Unfortunately, NSS is more for the server than the client.
As the browser becomes the desktop, it does not matter so much what Gtk and Qt can do. I would say that we should focus on AIR, MoonLight (as hated as it is), and HTML5 for 3D animations, streaming etc. From a security perspective, we might as well have a bunch of single purpose VMs making your hypervisor the closest thing to a desktop.
Of course, corporations like Intel can help. I would say that governments of all levels should be helping. The problem is that people/corporations don’t profit as much from free software. If they can add a little fee to each license they buy for the government, they make a profit doing nothing. The government is looking to get linux, OOo and other software for free. Where is the middle man going to make a living? As you can see, there are huge hurdles before any look and feel even enters the picture.
Personally, I think improving cross platform applications like firefox, OOo and the Mono upstarts (Banshee, etc) is the way to make the transition from proprietary OSes as easy as possible.
Cheers!
@Niko:
I’m not sure what you mean by ‘fail’, as the definition of a failure varies considerably. I not quite sure what you wanted to say in the remainder of the text; it’s badly formed english and the most likely interpretation makes it seem like trolling to me. Would you care to elaborate?
@LogicMagic:
The LSB is a great idea and the thought of a standard base is essential for the success of linux; sadly the project is too arrogant and ‘closed’ and seems to stray too much from the spirit of open source; look at all the fuss with Debian in the past, where in many occasions the Debian side was absolutely right to ignore the LSB and the LSB made no effort to listen to what Debian developers had to say.
I disagree about the importance of the underlying libraries [viz. GTK and Qt]. As the network becomes an inherent part of the computer, the OSes of the future will most certainly follow the model you suggest [albeit hopefully while still providing a great experience when the computer is offline]. This, however, does not mean that the underlying code shouldn’t be great; neither Silverlight/AIR/HTML5/JavaFX/whatever provide the performance and the same level of extremely rich experiences as native applications do; it’s a shame that the standard control design paradigm introduced by Microsoft and largely followed by everyone [save game developers and, perhaps, part of the Mac development community] doesn’t exactly allow for a proper illustration of this point today.
Thanks for the comments!
I thought about writing a lengthy response but I decided not to. I suggest however you reread my comment with this addition of information:
- I’m all for that corporations get behind Linux and have promoted this for a long time
- I’m not such a crack head that I believe in some myth about geeks hiding in basements writing code that run everything from TOP500 to Desktop Linux
- My living is administration of Windows servers and desktop, plus working directly with customers to meat their needs… yes I’m far from an ideologically driven student
- I didn’t even once say that Linux doesn’t need the support of a few strong corporations, I only said that it has the advantage of not being dependent on those corporations being the same ones all the time. Hence it might take a hit, get “hurt” but not fatal, find new partners and go on.
I don’t really know how you came to those conclusions about my opinions on several matters. Maybe this wouldn’t have happened if I had my own blog with enough entries to clarify my position. On the other hand I’m probably also mistaken about a couple of things about your views, so it’s probably fair and square. Certain remarks you made turn it into a semantic exercise. Why this talk about “it’s not something… there’s no computing idea” when you’re able to see beyond my maybe less eloquent wording.
Anyway I don’t mind. Let others read and decide for themselves. In my opinion all systems for the desktop of today are a mixed bag of pros, cons, gems and crap, including Linux. I’ve no problem in saying that Linux sucks, just as I can say about the others. Usually it comes down to what you need to achieve.
Just a last note: Yes Microsoft tries to sort out the kernel layers and define more exactly which calls go where. In my view this is a more fundamental issue and even with Microsoft’s economic resources a challenge. As you know Microsoft was stuck before one bright mind figured out a working approach. Yes, they’ve made progress but they’ve still a long way to go. Therefore you still have a mixed system where calls that should have gone to kernel base ends up in “old places”. I’m a bit puzzled to see you describe this lightly as something Microsoft is fixing like it’s already mission accomplished or close to. We actually don’t know if Microsoft is able to get it right; we only know that they’ve made some progress but it’s complex and there’s a long way to go. To slim down the kernel is still a delicate matter somewhat hazardous.
I hope you some day witness how your vision is fulfilled.
Hello,
@cosmix:
on KDE: From my point of view KDE evolves pretty fast and I feel it’s pretty usable right now, I’m using Kubuntu 9.04 and it’s good enough for me, But you are right with one aspect on it: most contributors to it are individual programmers – save for the Nepomuk project.
on innovation: But hasn’t always innovation come from small companies or individual contributors? Isn’t that how the GNU project started? In almost any aspect of the technical field innovation has come from a small startup/individual whose idea and technology was adopted by the mainstream. As they grew to make a big company and the fruits of their idea ripe, they lost capacity to innovate further. And that makes room for the next startup.
So in my opinion, the fact that innovation in technology fields comes from the small startup/individual contributors is actually the rule and not the exception.
What are the odds of two people thinking the same idea in approx. the same time?
Here http://linuxgeeksunited.blogspot.com/2009/05/big-boys-dont-innovate-anymore.html this guy makes a similar point, but he is more pessimistic – i.e. she/he doesn’t think big corp will become innovative again in our foreseeable future.
However, I still think the problem is elsewhere.
You say that “chaos, old and insufficient APIs and poor documentation are the norm in linux-land”. While I mostly agree that we are in a bad situation, I see things differently and a different way of moving forward,
I think corporations actually hold off the desktop because there is too much ego in desktop projects that they don’t work well together. Does anyone remember the crisis with the Gentoo Foundation and the project in general (roughly a year ago)? Read more on this here http://www.tectonic.co.za/?p=2030. It’s unlikely any company would want to invest in that project at that time. Aside from the need to make profit quickly, I think this kind of miscommunication is what is holding back corporations from the GNU/Linux desktop.
I see KDE as very innovative, as it brings Nepomuk and widgets to the desktop, and they do try to make documentation available through the http://techbase.kde.org/ and http://userbase.kde.org/. Gnome have started their own re-engineering process for Gnome 3.0, which hopefully will produce a more consistent API and documentation. Also, KDE and Gnome hold Akademy and Guadec together (see http://grancanariadesktopsummit.org/) which will hopefully produce a more consistent API/interface. Also LSB from Linux Foundation is growing stronger.
So even if there is a gap in communication inside a project (I know there is miscommunication inside KDE because I follow KDE Planet regularly) and outside (see the Gnome vs. KDE debate) the situation is getting better because there are conscious efforts to tighten this gap.
Financially-wise, as profits from proprietary platforms decrease or stagnate, more and more corporations will see GNU/Linux as the next platform for earning money. Also corporations are like a herd: once one has broken the psychological barrier of using Linux as a (primary) platform others tend to follow the lead. And the timing couldn’t be more right.
PS: I mostly gave examples from the KDE because that project I know best.
@Bogdan, thanks for those two great comments. I don’t think the two posts are that similar, neither that the idea is unique or difficult to come up with. Many people are concerned about the future of linux on the desktop.
I think KDE shows promise, in the same way KHTML showed promise in 2001 and 2002; but the promise became reality only when Apple took it and made Webkit; KHTML is still horribly broken for too many sites while Webkit is a fantastic engine — arguably the best one around from a technical standpoint. At the same time KHTML is still lagging behind both Gecko and Webkit.
I disagree with you that KDE is mature; innovation is a multi pronged challenge and KDE is nothing but a soup of pretty interesting ideas and some implementations; I love some aspects of it, but it’s by no means finished and in no way well thought out in terms of usability, maturity or performance compared to, say, OS X. On my linux workstation KDE 4.2.2 performs poorly even compared to GNOME: its very unstable, it is clearly unfinished in some areas, it has lousy support for dual head displays, it has ridiculous issues with sound and — perhaps above all — it lacks useful, good native applications.
Widgets on the desktop are nothing new; they’ve been around for more than 15 years in some form or another, even though they only took off in the last 5. Nepomuk has some interesting concepts, but it’s too early to tell whether it’ll live up to expectations; and mind you I’m not really interested in discussing what’s good in KDE and what’s not for hackers, developers, geeks or enthousiasts. Those are only 2%. I’m talking about a polished, usable desktop by the masses that is easy to use and at the same powerful. KDE is not there; KDE probably won’t be there in the next 3-4 years. GNOME is even further behind, exactly because corporations like Redhat, Novell etc. have been solely interested in the enterprise desktop — a dumbed down terminal where people can access email, access the web and that’s about it.
You’re correct that there are conscious efforts to make the linux desktop better; freedesktop.org is one of them. Communication between large projects is another. Cleaning up the cruft that’s been there for years is yet another. But in all those cases there’s very little focus, and very little vision and very little agreement. And most of the time we end up reinventing the wheel instead of building stuff on good foundations.
Large corporations are seldom innovative by themselves. Large corporations that fund or buy small innovative startups are. After more than 10 years of using Open Source I’ve come to this conclusion [as sad as it may be]:
Totally open design by committee is great in theory, but very hard in practice. Unless you can provide an open, timely and efficient workflow process for developers, designers, architects etc. to create something you’re either going to be very late in delivering, very bad in engineering/design/etc. or will just fail.
Focus is required above all when it comes to the linux desktop and neither project [KDE or GNOME] nor anyone else right now has this. KDE will most certainly get better in the next few years and is showing a lot of promise. GNOME 2.30/3.x will most probably kickstart a new cycle for that project too, although it’s not clear where that will lead the project right now. But unless we come to think of the desktop as one thing, unless we take a holistic approach of the whole software stack, we’re always going to be behind the leaders.
What the linux desktop needs in my opinion is some large corporation (say someone like IBM) to fund a number of relatively small groups of excellent developers, usability experts, graphic designers, documentation writers, etc., task them to come up with a great, cohesive desktop platform that uses good code that’s been out there until now and creates something new, and let them do the work. The large corporation could donate the code to a Foundation, but still retain the copyright for its own version [that it might sell].
In a linux kernel-like paradigm, the acceptance of patches for that distribution should be controlled by some [committee], but anyone would be capable of getting the code, seeing it, forking it etc.
This would be both [somewhat] open and more efficient than the proposition that KDE and GNOME are offering because it offers focus and a great start. Good documentation, clear objectives, a meaningful roadmap [not littered by pointless 'games' or useless 'utilities'].
I agree this is a possibility and would very much like to see it happen.